It goes around in a big sticky circle, doesn’t it. I believe the thinking may go something like this: I ate pretty much what the scientists (or the marketing execs) said I should eat all these years and look where it got me. So obviously weight gain isn’t science, it’s some kind of magic. So who knows, maybe only eating potatoes is the secret magic that will work.
Disturbingly, it sounds like your friend’s doctor swallowed the same blue pill.
More or less anyone who goes on a diet will lose some weight initially, especially if the magic food is a single item and not an industrial concoction designed to be hyper palatable. It doesn’t matter if it’s potatoes or celery or grapefruit or bacon. Even just keeping a food diary usually results in weight loss - when you think that much about what you’re eating, you eat less.
The hard part is not changing the food from potatoes to keto friendly food, it’s changing the behavior from seeking magic to seeking science. Which is also a conundrum: how do you create a big sexy non-scientific soundbite, in order to encourage comparatively long, complex scientific thinking.
Do you know if there’s a way to come in from the side? That is, to target something other than the obesity? For example, if he’s diabetic, he probably knows sugar isn’t good for him. If he learns that potatoes are basically sugar in a different package, maybe he’ll see that the potato diet is a bad idea at least for him personally, and be more open to a different idea?